通信学报 ›› 2020, Vol. 41 ›› Issue (9): 8-20.doi: 10.11959/j.issn.1000-436x.2020200

• 专题:面向智慧应急的通信与计算融合 • 上一篇    下一篇

基于深度强化学习的应急物联网切片资源预留算法

孙国林1,欧睿杰1,刘贵松1,2   

  1. 1 电子科技大学计算机科学与工程学院,四川 成都 611731
    2 电子科技大学中山学院,广东 中山 528402
  • 修回日期:2020-08-19 出版日期:2020-09-25 发布日期:2020-10-12
  • 作者简介:孙国林(1978- ),男,河北唐山人,博士,电子科技大学副教授、硕士生导师,主要研究方向为人工智能、区块链和移动智能系统等|欧睿杰(1989- ),男,四川成都人,电子科技大学博士生,主要研究方向为人工智能、区块链、移动网络资源管理等|刘贵松(1973- ),男,山东临沂人,博士,电子科技大学教授、博士生导师,主要研究方向为类脑计算、机器学习和模式识别与智能系统等
  • 基金资助:
    国家自然科学基金资助项目(61771098);四川省科技计划基金资助项目(2020YFQ0025)

Deep reinforcement learning-based resource reservation algorithm for emergency Internet-of-things slice

Guolin SUN1,Ruijie OU1,Guisong LIU1,2   

  1. 1 School of Computer Science and Engineering,University of Electronic Science and Technology of China,Chengdu 611731,China
    2 Zhongshan Institute,University of Electronic Science and Technology of China,Zhongshan 528402,China
  • Revised:2020-08-19 Online:2020-09-25 Published:2020-10-12
  • Supported by:
    The National Natural Science Foundation of China(61771098);The Science and Technology Research Program of Sichuan Province(2020YFQ0025)

摘要:

针对应急物联网(EIoT)超低时延服务需求,设计了面向超低时延传输应急物联网的多切片网络架构,提出 EIoT 切片资源预留和多异构切片资源共享与隔离的方法框架。所提框架采用深度强化学习方法实现实时异构切片间资源需求的自动预测与分配,切片内用户资源分配建模为基于形状的二维背包问题并采用启发式算法数值求解,从而实现切片内资源定制化。仿真结果表明,基于资源预留的方法能够使 EIoT 切片显式保留资源,提供了更好的安全隔离级别;深度强化学习能够保证资源预留的准确和实时更新,有效兼顾资源利用率和切片差异化服务质量要求。与4个已有算法对比表明,Dueling DQN具有更好的性能优势。

关键词: 应急物联网, 深度强化学习, 资源预留, 超低时延通信

Abstract:

Based on the requirements of ultra-low latency services for emergency Internet-of-things (EIoT) applications,a multi-slice network architecture for ultra-low latency emergency IoT was designed,and a general methodology framework based on resource reservation,sharing and isolation for multiple slices was proposed.In the proposed framework,real-time and automatic inter-slice resource demand prediction and allocation were realized based on deep reinforcement learning (DRL),while intra-slice user resource allocation was modeled as a shape-based 2-dimension packing problem and solved with a heuristic numerical algorithm,so that intra-slice resource customization was achieved.Simulation results show that the resource reservation-based method enable EIoT slices to explicitly reserve resources,provide a better security isolation level,and DRL could guarantee accuracy and real-time updates of resource reservations.Compared with four existing algorithms,dueling deep Q-network (DQN) performes better than the benchmarks.

Key words: emergency IoT, deep reinforcement learning, resource reservation, ultra-low latency communication

中图分类号: 

No Suggested Reading articles found!